New ARES framework uses GNNs and Half-Space Trees for edge anomaly detection

Researchers have developed ARES, an unsupervised framework designed to detect anomalies in streaming temporal graphs. This model combines Graph Neural Networks (GNNs) for feature extraction with Half-Space Trees (HST) for anomaly scoring, enabling it to identify unusual temporal connections in real-time. ARES addresses challenges like concept drift and large data volumes by embedding node and edge properties to capture anomalous behaviors. The framework can also incorporate a simple supervised thresholding mechanism using minimal labeled data to adapt to different domains. AI

IMPACT This research could improve real-time threat detection in dynamic network environments.

RANK_REASON The cluster describes a new research paper detailing a novel model for anomaly detection in streaming graphs. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• alphaXiv
• arXiv
• CatalyzeX
• DagsHub
• Gotit.pub
• Graph Neural Networks
• Half-Space Trees
• Hugging Face
• IArxiv
• ScienceCast
• Simone Mungari

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →